The chi-square test for independence

Abstract

This chapter will look at the most popular technique for hypothesis testing of categorical data. This is the chi-square test of independence. The earlier chapters emphasized that the choice of inference test is mainly affected by two considerations:

• the number of samples to be compared (i.e. one sample, two samples, or more than two samples);

• the level of measurement of the variable with which the samples are to be compared (i.e. nominal, ordinal, or interval/ratio).